The overall objective of this project is to develop strategies, which will allow the efficient and safe treatment of hematopoietic diseases by hematopoietic stem cell (HSC) gene therapy. Unfortunately neither in vitro studies nor studies in the mouse model have been predictive for safety and efficacy of gene transfer in large animals or humans, thus we will use the dog model which will also allow us to evaluate improved strategies directly in a disease model. While gene transfer efficiencies into HSCs in the dog model and other large animals have improved significantly over the past several years, the development of leukemia in 3 patients in the French X-linked severe combined immunodeficiency (XSCID) gene therapy trial has shifted the emphasis from efficacy to safety. Thus, we will use the dog model to study the safety of 3 commonly used integrating vector systems: gammaretrovirus, lentivirus, and foamy virus vectors. Using these vector systems, we have been able to achieve efficient transduction of canine long-term repopulating cells with stable gene transfer levels >5% in a significant number of dogs. Thus, we now have a unique resource available to study in Specific Aim 1 the safety of HSC gene transfer with these vectors. In Specific Aim 2, we will further optimize transduction conditions for foamy and lentiviral vectors to minimize risks from insertional mutagenesis. We will focus on lentivirus and foamy virus vectors since, in contrast to gammaretroviral vectors, these vector systems allow for efficient transduction using short transduction cultures. This is particularly important for stem cell gene therapy in a nonmyeloablative transplant setting where maintenance of stem cells is crucial for the ability to compete with surviving endogenous stem cells. Thus, in Specific Aim 3 we will use these vector systems in a nonmyeloablative setting and also explore whether in vivo selection strategies can improve gene transfer levels after nonmyeloablative conditioning. Finally, in Specific Aim 4 we will test improved gene transfer protocols in a canine genetic disease model. The availability of a clinically relevant large animal model should allow us to quickly translate our findings to clinical HSC gene therapy studies.